jarin udom's tumblelog

1. Use an extremely restrictive and viral license that will force companies and “serious” developers to release the source code of derivative works. You can forget about that commercially viable fork of Diaspora that you wanted to make.
   [[posterous-content:pid___0]]
2. Require open source contributors to sign a contract giving them joint ownership of any contributions you make. Of course, you don’t get joint ownership of or any rights to any derivative works they make using your work. That would just be unfair, right?
   [[posterous-content:pid___1]][[posterous-content:pid___2]] 

In order for Diaspora to compete with Facebook, it’s going to need massive and rapid developer adoption, not just from Free Software enthusiasts but from companies and financially-motivated developers. Imagine how Diaspora would blow on to the scene if Coca-Cola created their own branded Diaspora seed (but compatible with all of the other Diaspora seeds out there) with their own music discovery features. Or if Yahoo! created one for their 622 million users with deep integration with Yahoo! mail and Flickr. These are features that belong on their own forks and not bloating up the main codebase, but there’s no way that Coke or Yahoo! would ever create these if they had to publish the entire source code.

Sure, the anti-corporate crowd is probably saying “That’s the point! Down with big business! Power to the people!”, but in order for a new social network protocol to compete with Facebook it needs users. Lots of them. Hundreds of millions of them. And guess what happens when companies massively adopt open source software? They contribute back to those projects and make them more secure, more useful, and more stable. Just look at WebKit, SproutCore, MySQL, Apache, and any number of other commercially-used open source projects.

“But corporations will destroy your privacy and lock you in!” The Diaspora project was created entirely in response to Facebook’s terrible views on privacy and data portability. Guess what? The Diaspora protocol makes it easy to move your profile data to another seed. But if there aren’t any better seeds out there, what’s the point? Sure, you can create your own seed and move all of your data over to it, but how many non-developers even know how to clone a git repository, let alone install MongoDB and deploy a web application?

For Diaspora to succeed, it needs developers and/or companies to create larger seeds that regular mortals can just sign up for and start posting their baby photos. Those things don’t come cheap, though. Facebook spends over $2 million per month on hosting images alone. Developers and companies need the freedom to make modifications to Diaspora to help pay those costs without having to reveal the “secret sauce” bits to the world (many of which would have absolutely no use to the open source community anyway).

I was pretty excited when the developers released the Diaspora source code yesterday. I even stayed up late hacking away at it last night to make it work with Heroku, so that anyone who wanted to would be able to have their own personal Diaspora instance without having to pay any hosting costs. Will my changes get merged into the main codebase so that they will be more accessible to everyone? Probably not, because there is no way in hell I am signing a contract just to contribute to an open source project.

I think the Diaspora dev team is well-intentioned but naive. I just wish they had used that part of the $200,000 they raised from the community on things other than paying lawyers to sabotage their vision before it even got off the ground. The only things that can save Diaspora from just being a social network for Free Software and privacy nuts now are for the developers to publish the protocol documentation and change the reference implementation license to something less restrictive (like the BSD license).

Posted via email from Jarinheit | Comment »